PureThermal 1 firmware v1.0.0 preview (Lepton 3 and more)

629 views
Skip to first unread message

Kurt Kiefer

unread,
Sep 10, 2017, 11:09:28 PM9/10/17
to Flir Lepton
Hi everyone,

This is a heads up that some new code has been pushed on the development branch of the PT1 reference firmware on github (https://github.com/groupgets/purethermal1-firmware/tree/development). Major changes:

1. Support for Lepton 3, same firmware image for Lepton 2 (sensor type determined when power applied)
2. No more special Y16 firmware, acquiring this data type dynamically disables AGC and RGB and returns raw 14-bit data
3. Reduce the jitter and dropped frames during acquisition, as well as lower latency. Streams full-rate (~26 fps) on Lepton 2, including duplicate frames.
4. Drop 12-bit YUV (i420/YV12) formats (some camera programs would choose them by default over 422 or RGB formats)

There have been a lot of people asking for updates for Lepton 3, so I wanted to post this here to give people a chance to try it and provide any feedback before stamping an official release. This firmware will work for F411 and F412 boards. Give it a try and let us know how it's working for you!

Kurt

Clive

unread,
Apr 3, 2018, 3:48:46 AM4/3/18
to Flir Lepton
Hi Kurt

Just got my new 3.5 Radiometric on PT2. Can't find documentation on default firmware. Will the below link work with 3.5?

Thanks

Clive 

Kurt Kiefer

unread,
Apr 3, 2018, 12:29:35 PM4/3/18
to Flir Lepton
PT2 firmware is currently on a separate branch: https://github.com/groupgets/purethermal1-firmware/tree/development+pt2

The firmware is almost exactly the same as the PT1, except for the different 8MHz clock used so that USB DFU works.

The plan is to do some board detection before setting up the clocks and the PT1 and PT2 will eventually share a code base.

Clive

unread,
Apr 4, 2018, 5:44:57 PM4/4/18
to Flir Lepton
Thanks Kurt. 

The flash instructions in readme remain similar to the PT1? Is there some change for direct flash over USB?

Cheers 

Kurt Kiefer

unread,
Apr 4, 2018, 5:52:01 PM4/4/18
to Flir Lepton
Good question.

If you have a pogo-pin jtag adapter (e.g. the one included with the pt2 pro kit) then the flash instructions are the same.

JTAG is no longer required for the PT2 though. Reaching into the wayback machine on github, you can find USB DFU instructions as for very early PT1 boards.

Anyhow thanks for pointing this out, we'll get this documented properly.

Clive

unread,
Apr 11, 2018, 4:56:17 AM4/11/18
to Flir Lepton
Thanks Kurt 
I want to capture an image from the PT2 / Lepton 3.5 and have the raw values of the radiometric but it is a quagmire as to how to do this. I am using python and I assume that capturing the image should be possible with https://github.com/groupgets/purethermal1-uvc-capture and the uvc-radiometry.py but this only states it works for Lepton 2.5. I have been buying FLIR's / PT1 / PT2 and still cannot get a simple raw (or a radiometric) thermal image. Any hints.
Thanks 

Kurt Kiefer

unread,
Apr 11, 2018, 11:40:19 AM4/11/18
to Flir Lepton
That example should work; it was written before the 3.5 existed, hence the note. I think you should just be able to change the frame size:


GetThermal should also give you radimetric imaging:

Clive

unread,
Apr 16, 2018, 3:30:22 AM4/16/18
to Flir Lepton
Kurt

I am generally use windows and have been delving into Ubuntu to get the raw image output into python but after three days I cannot get any further than
ERROR: pipeline could not be constructed: no element "v4l2src".
I have PT1 and PT2 boards and Leptons 3 / 3.5 but have not been able to extract the raw data no matter what I try. Any more tips?

Cheers

Joshua Ingham

unread,
Jun 5, 2018, 7:36:39 AM6/5/18
to Flir Lepton
In there any update upon this?

Kurt Kiefer

unread,
Jun 5, 2018, 12:03:35 PM6/5/18
to Flir Lepton
Considering that the uvc_radiometry example does not use v4l2 (it uses libuvc), that error message does not make any sense.

Since this thread was posted, that example was updated to automatically choose the different frame size of the Lepton 3.5 if necessary, so it might make sense to give it another try after pulling down the latest code.

Kurt
Reply all
Reply to author
Forward
0 new messages